all_results.json

{
    "epoch": 4.988190836088805,
    "eval_dpo_loss": 114.26480102539062,
    "eval_logits": -0.652398407459259,
    "eval_logps": -81.85186004638672,
    "eval_loss": 223.14205932617188,
    "eval_objective": 224.39108276367188,
    "eval_ranking_idealized": 0.5092975497245789,
    "eval_ranking_idealized_expo": 0.5092975497245789,
    "eval_ranking_simple": 0.5082644820213318,
    "eval_regularize": 224.39108276367188,
    "eval_runtime": 260.3381,
    "eval_samples": 5790,
    "eval_samples_per_second": 22.24,
    "eval_steps_per_second": 0.93,
    "total_flos": 0.0,
    "train_loss": 74.13437041504817,
    "train_runtime": 35293.9194,
    "train_samples": 50802,
    "train_samples_per_second": 7.197,
    "train_steps_per_second": 0.025
}